Wasted ad spend hides in four layers: what you bid on, where the click lands, whether the path works, and whether you are measuring it correctly. The last one is the most dangerous, because it quietly corrupts the other three.
Almost every paid search account wastes money. That is not an accusation; it is the nature of auction-based advertising run at scale, where search behaviour shifts daily and small misconfigurations compound quietly over months.
The useful questions are how much, and where.
How much is typical?
Published estimates vary, and it is worth being honest about their limits. A frequently cited WordStream study of around 500 small and mid-sized Google Ads accounts found that the average advertiser was wasting about a quarter of its budget. Seer Interactive has reported companies wasting around 15% of spend on irrelevant keywords alone. Both skew towards smaller accounts, and the WordStream data is more than a decade old, so treat them as indicative rather than as a benchmark for your account.
The range we plan around is 10–30% of spend. Well-run accounts sit at the lower end. Accounts nobody has looked at closely in a year tend to sit at the upper end or above it.
More useful than the headline number is knowing where the waste accumulates. It tends to collect in four layers.
Layer 1: Keywords and match types
What it is: paying for searches that are too far from your offer to ever convert.
The usual culprit is broad match drift. Broad match lets the platform show your ad for searches it considers related, and it interprets “related” generously. A campaign built for a specific B2B service gradually starts buying searches from job seekers, students, people looking for free templates, and buyers of adjacent products you do not sell.
How to check: open the search terms report — the actual queries that triggered your ads, not the keywords you chose — sorted by spend. Read the top of it line by line. Anything that could never become a customer is waste, and it should become a negative keyword.
Layer 2: Intent alignment
What it is: high-intent searches landing on generic pages.
Someone searches for a specific, purchase-ready phrase and lands on your homepage, or a broad services page, and has to hunt for what they came for. Many leave. This is the stages of awareness problem showing up in the ad account.
It costs you twice. You lose the conversion, and you usually pay more for the click. Google evaluates expected clickthrough rate, ad relevance and landing page experience in real time as part of each auction; weaker relevance generally means paying more to hold the same position.
How to check: for your highest-spending ad groups, compare the search intent with the page it lands on. If the page does not answer the searcher’s question in the first screen, it is misaligned.
Layer 3: User paths
What it is: the click worked, and then the experience broke.
The landing page returns an error. The product is out of stock or discontinued. The page loads slowly on mobile. The form is broken, or asks for twelve fields. The phone number is wrong. These are not strategy problems; they are maintenance problems. They are also surprisingly common, because nobody clicks their own ads.
How to check: crawl every final URL in the account for errors. Then go through the path yourself on a phone — click, land, and try to convert — for your top campaigns.
Layer 4: Tracking integrity — the quiet killer
What it is: measuring the wrong thing, or measuring the right thing wrongly.
- Conversions counted twice, because a tag fires on page load and again on submit
- Micro-actions like page views or button clicks set as primary conversions alongside real leads
- Offline revenue invisible to the platform, so it cannot tell a junk lead from a closed deal
- Cross-domain gaps that lose the source of the conversion entirely
This layer is the most dangerous for a specific reason: it corrupts the other three. If your tracking is wrong, you cannot see waste in the other layers accurately, and any fix you make is measured against a broken yardstick.
Worse, modern ad platforms bid automatically against whatever you tell them a conversion is. If that signal is wrong, the algorithm becomes confidently wrong at scale — and automation makes it wrong faster. This is audit before you automate in its most concrete form.
What to fix first
- Tracking. Nothing else can be measured reliably until this is sound. Audit your conversion actions, remove duplicates, and decide what counts as a real conversion.
- User paths. The cheapest fixes in the account. A broken page repaired this afternoon stops the loss immediately.
- Intent alignment. Route high-intent searches to pages that match them.
- Keywords. Add negatives, tighten match types, and put the search terms review on a schedule — because the waste comes back within a quarter if nobody watches it.
Where to go next
The Google Ads audit guide is the step-by-step version of this. Attribution threading covers fixing tracking at the joints between systems. And optimise on revenue, not activity explains how to point bidding at revenue once the tracking can be trusted.